Comments and Discussion

Article excerpt

COMMENT BY BARRY EICHENGREEN In this paper Niall Ferguson presents us with an interesting paradox. Despite the fact that the United States has been fighting an expensive war on terror in Iraq, Afghanistan, and other countries, and despite an increased awareness since 9/11 of the possibility of another catastrophic attack on a major U.S. city, U.S. asset markets delivered healthy returns in the five years following the attacks on the World Trade Center and the Pentagon.

There are three possible explanations for the Ferguson paradox. First, it could be that the kind of catastrophic event that provides the motivation for the paper is actually a low-probability event and is accurately perceived as such. To be sure, the author cites an estimate by Matthew Bunn putting the odds of a nuclear terrorist attack over the next ten years at 29 percent. These odds are obtained as the product of the probability that terrorist groups will attempt to secure the components of a nuclear weapon in a given year times the probability that they will succeed times the probability that they will devise a workable nuclear device times the probability that they will successfully detonate it in a major population center; the resulting probability is then multiplied by the posited number of terrorist groups and years. But the numerical values assigned to these parameters are arbitrary. The exercise is data free. The limits of calibration as a guide to policy will be familiar to the readers of this journal. And whatever one thinks about the value of calibration exercises in macroeconomics, we probably know more about how to calibrate the intertemporal elasticity of substitution than about the probability that a terrorist group will be able to assemble a workable nuclear device out of its components.

Second, even if there is a nonnegligible probability of that catastrophic event occurring over the horizon relevant to investors, the latter may display disaster myopia. As formulated by Jack Guttentag and Richard Herring and applied to international banking, (1) this thesis is that cognitive bias leads lenders to excessively discount the probability of a significant negative shock or disaster as a function of the length of time since such a shock last occurred. (2) Even if only some lenders suffer from disaster myopia, if the market is competitive, other lenders may feel pressure to go along in order to avoid losing market share. Hence the risk of a relatively rare catastrophic event will not be fully reflected in loan prices and conditions. There is an obvious sense in which recent events lend support to the notion that financial markets can neglect the possibility of extreme realizations. As someone who has invested heavily in real estate squarely atop one of the world's most active earthquake faults, I take the hypothesis of disaster myopia seriously (introspection being the obvious alternative to calibration as a substitute for empirical work).

Third, even if the probability of a catastrophic event is nonnegligible, and even if investors perceive this accurately, the impact on returns may be swamped by other factors. The Federal Reserve cut interest rates sharply in 2001 and reversed out those cuts only gradually. U.S. financial markets thus enjoyed an accommodating environment through the first half of the decade. Low interest rates raise the capitalized value of expected future profits. An elastic supply of credit supports purchases of financial assets. Now that the bubble has burst, there is a growing awareness of how Federal Reserve policy contributed to the asset market inflation of this period, not just in real estate but also in the securities markets that are Ferguson's focus. Permissive financial regulation permitted banks and nonbank financial institutions to increase their leverage and expand their balance sheets. Foreign central banks, seeking to prevent their currencies from appreciating and anxious to augment their reserves, generously financed the U. …

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.